Based on Gaussian filter to improve the effect of the images in Gaussian noise and pepper noise

Abstract

Abstract Gaussian filter is one of the important research topics. Researchers find that Gaussian filter can suppress noise, but there is still a gap in the completeness of Gaussian filter for denoising. Therefore, the author optimized the Gaussian filter to achieve better filtering effect. By collecting SNR and PSNR data and comparing and analysing different noise types of the data under the same condition, the author explored the denoising effect of Gaussian filter method on Gaussian noise and Salt and pepper noise, and improved the Gaussian filter method. According to the comparison of SNR and PSNR data, the author found that under the default condition, the SNR value of Gaussian noise after optimization is 57.69201, and the SNR value of Salt and pepper noise is 31.5896. The PSNR value of Gaussian noise is 31.5021, and the PSNR value of Salt and pepper noise is 19.7872. The SNR and PSNR values of Gaussian noise are much larger than those of Salt and pepper noise, indicating that the denoising effect of Gaussian filter on Gaussian noise is obviously better than that of Salt and pepper noise under the same condition. The SNR value and PSNR value of Gaussian noise in the two groups of experiments are larger than Salt and pepper noise, which indicates that under the same condition, Gaussian filter is conducive to suppressing Gaussian noise and has a more significant impact on Gaussian noise.